LIVE · refreshes every 20 min

updated Nov 14, 07:23 UTC

110010.art / archive

Research

Nov 12

2d ago MIT researchers develop technique to streamline creation of fast, reliable, and adaptable AI models for climate prediction MIT News (AI) → story

Nov 11

3d ago FaST: Factored Synthesis of Transitions for Offline Reinforcement Learning arXiv cs.LG → story

3d ago Framing-induced vulnerability in large language models: evidence from Chinese news headlines arXiv cs.CL → story

3d ago GRAP: Gender Rubric-Aware Probing for Analyzing Gender Bias in LLMs arXiv cs.AI → story

3d ago Graph visualizations in prompts for knowledge graph question answering arXiv cs.AI → story

3d ago HiCAP: hierarchical cultural adaptation via personality for dialogue generation arXiv cs.CL → story

3d ago High probability risk control for online policy learning arXiv cs.LG → story

3d ago IC-GNN: modeling individualized brain community organization for disease prediction arXiv cs.LG → story

3d ago IIT Patna participates in ICPR 2026 TVRID Competition on Privacy-Preserving Person Re-Identification from Top-View RGB-Depth Camera using 86 identities across four RealSense D455 cameras arXiv cs.AI → story

3d ago India's languages incur higher token costs in 2026 due to tokenizer disparities across nine public tokenizers, per arXiv 2611.13058; 25 Indian languages in FLORES-200 analyzed. arXiv cs.CL → story

3d ago IntentTwin: Deconfounding candidate support and the oracle-to-deployment gap in offline web behavior prediction arXiv cs.CL → story

3d ago Is one temperature enough? Inference-time sampling values and when temperature diversity helps in pass@K arXiv cs.LG → story

3d ago Khammoon: From a linguistic guideline to a Thai morpheme segmenter arXiv cs.CL → story

3d ago LandslideFM: a foundation model for unified landslide perception and understanding arXiv cs.AI → story

3d ago Linear probes to Bayesian manifolds: geometry of arithmetic in large language models arXiv cs.LG → story

3d ago MADSNet: a multi-scale axial-attentive dense spiking network for spike camera image reconstruction. arXiv cs.LG → story

3d ago Measuring definition sensitivity in LLM annotation of social science concepts arXiv cs.CL → story

3d ago Mobile AI Factory builds a fine-tuned, quantized small language model for a given task and mobile device. arXiv cs.CL → story

3d ago Monitoring a visible channel alters learned communication; adaptive reallocation of joint message-route codes observed in end-to-end game experiments arXiv cs.AI → story

3d ago Multimodal attributed graphs are surveyed across models, benchmarks, and frontiers, covering graph-centric and foundation-model-centric approaches to learning on MAGs. arXiv cs.LG → story

3d ago One Table Benchmark evaluates generalist tabular models across generation, imputation, and prediction tasks across 56 datasets and 46 methods arXiv cs.LG → story

3d ago Permutation-based pre-trained model merging via reciprocal orthogonality arXiv cs.LG → story

3d ago Physics2Weight: direct neural network architecture and weight generation from physical laws using large language models arXiv cs.AI → story

3d ago Post-training influences verbalized algorithmic computation in language models; probing shows carry information remains recoverable from internal activations across SFT and RLVR checks arXiv cs.CL → story

3d ago Prediction-powered LLM ranking with adaptively collected human feedback arXiv cs.LG → story

3d ago Predictive surprise as self-grounding concept bottleneck for interpretable time series arXiv cs.LG → story

3d ago RadRef: a reference-free metric for radiology report generation arXiv cs.LG → story

3d ago Reinforcement learning-based control of cancer chemotherapy treatments arXiv cs.AI → story

3d ago Rethinking in Spikes: mitigating hallucinations in MDLMs with step-aware decoding arXiv cs.CL → story

3d ago Seeing is not believing: counterfactual lesion erasure teaches a medical vision-language agent to report only what the image shows arXiv cs.AI → story

3d ago SetPrint: Black-box LLM fingerprinting via setwise correspondence of response changes arXiv cs.AI → story

3d ago Shape of a program: path signatures for trace-to-program induction arXiv cs.LG → story

3d ago Symbolic Imitation Learning distills an oracle policy into a decision list with linear model motion primitives using offline demonstrations and online tuning. arXiv cs.LG → story

3d ago Third Eye: immediate gains from self-update are near zero; further native updates can improve performance, according to tests on 3-8B backbones across math and code arXiv cs.LG → story

3d ago TimeOmni-o1: bidirectionally coupling time series foundation models and LLMs for multimodal reasoning and forecasting arXiv cs.CL → story

3d ago Train on prefill, monitor at decode: TopicScope trains causal topic classifiers on SAE features to label every token and allow topic changes mid-response arXiv cs.AI → story

3d ago Understanding students' rationales for AI-mediated cheating: an instructor perspective from online forums analysis arXiv cs.CL → story

3d ago Understanding-first AI engineering: purposes, reference architecture, and architectural patterns for human understanding arXiv cs.AI → story

3d ago Which successes make good references? Action-sequence alignment for multi-turn on-policy distillation arXiv cs.LG → story

3d ago ZeroPoint: Lightweight energy-based decision model arXiv cs.AI → story

←12 3…30 →

Older items: monthly archive →